Giter Club home page Giter Club logo

gpv's Issues

Train

Hello, do you have any plans to share training part?

ModuleNotFoundError: No module named 'numba.decorators'

Hello,
I am trying to run GPV on my Linux.
I use venv. Here is my pip list

(project_env) ds@ds-Standard-PC-i440FX-PIIX-1996:~/GPV$ pip list
DEPRECATION: The default format will switch to columns in the future. You can us                                                                                                             e --format=(legacy|columns) (or define a format=(legacy|columns) in your pip.con                                                                                                             f under the [list] section) to disable this warning.
aiocontextvars (0.2.2)
audioread (2.1.8)
certifi (2020.6.20)
cffi (1.14.1)
chardet (3.0.4)
contextvars (2.4)
cycler (0.10.0)
dcase-util (0.2.16)
decorator (4.4.2)
future (0.18.2)
idna (2.10)
immutables (0.14)
joblib (0.16.0)
kiwisolver (1.2.0)
librosa (0.7.1)
llvmlite (0.33.0)
loguru (0.4.0)
matplotlib (3.3.0)
numba (0.50.1)
numpy (1.18.2)
pandas (1.0.3)
Pillow (7.2.0)
pip (9.0.1)
pkg-resources (0.0.0)
pycparser (2.20)
pydot-ng (2.0.0)
pyparsing (2.4.7)
PySoundFile (0.9.0.post1)
python-dateutil (2.8.1)
python-magic (0.4.18)
pytz (2020.1)
PyYAML (5.3.1)
requests (2.24.0)
resampy (0.2.2)
scikit-learn (0.22.2.post1)
scipy (1.4.1)
sed-eval (0.2.1)
setuptools (39.0.1)
six (1.14.0)
SoundFile (0.10.3.post1)
torch (1.5.0)
tqdm (4.43.0)
urllib3 (1.25.10)
validators (0.17.1)

This error I've got when launched the program

(project_env) ds@ds-Standard-PC-i440FX-PIIX-1996:~/GPV$ python3 forward.py -w so                                                                                                             und.mp3
Traceback (most recent call last):
 File "forward.py", line 10, in <module>
   import librosa
 File "/home/ds/GPV/project_env/lib/python3.6/site-packages/librosa/__init__.py                                                                                                             ", line 12, in <module>
   from . import core
 File "/home/ds/GPV/project_env/lib/python3.6/site-packages/librosa/core/__init                                                                                                             __.py", line 123, in <module>
   from .time_frequency import *  # pylint: disable=wildcard-import
 File "/home/ds/GPV/project_env/lib/python3.6/site-packages/librosa/core/time_f                                                                                                             requency.py", line 11, in <module>
   from ..util.exceptions import ParameterError
 File "/home/ds/GPV/project_env/lib/python3.6/site-packages/librosa/util/__init                                                                                                             __.py", line 77, in <module>
   from .utils import *  # pylint: disable=wildcard-import
 File "/home/ds/GPV/project_env/lib/python3.6/site-packages/librosa/util/utils.                                                                                                             py", line 15, in <module>
   from .decorators import deprecated
 File "/home/ds/GPV/project_env/lib/python3.6/site-packages/librosa/util/decora                                                                                                             tors.py", line 9, in <module>
   from numba.decorators import jit as optional_jit
ModuleNotFoundError: No module named 'numba.decorators'

I found that Numba will remove the shim for numba.decorators.jit at 0.50 here so I tried to changed numba in requirements to numba==0.48 and also changed torch to torch==1.5.0, because 1.4.1 was not found. After this It seems all works.

Improving Performance on Shorter Audio Clips

Using your GPVAD/VADC, I wish to process smaller chunks (i.e. ~200ms chunks) of audio files. However, when the duration is this low, the performance of the VAD is poor. What can I do to better the performance? I assume this must be done in the training side. Would you recommend downloading the datasets and splicing them into these smaller chunks, retraining from scratch?

Curious to hear your thoughts. Thank you!

bug in feature extraction

Hi

Extracting mel spectrogram features (line 38 at forward.py), you get the mel of resampled signal with the original sampling rate (sample_rate) instead of SAMPLE_RATE.

process the audioset

Thanks for your hard work! I wrote a train.py to tain the network, but I can't get the same results as your pretrained model provided, I wonder if something wrong with my audioset processing, Could you share your processing audioset code or train.py?

Thoughts on streaming the forward pass?

This is some really good work!
I have a question: Have you tried using your algorithm to process an audio stream? How would performance be affected? And how feasible would real-time processing be?

Thanks!

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.